CN109670224A - A kind of multirobot service border circular areas searching algorithm - Google Patents
A kind of multirobot service border circular areas searching algorithm Download PDFInfo
- Publication number
- CN109670224A CN109670224A CN201811492950.5A CN201811492950A CN109670224A CN 109670224 A CN109670224 A CN 109670224A CN 201811492950 A CN201811492950 A CN 201811492950A CN 109670224 A CN109670224 A CN 109670224A
- Authority
- CN
- China
- Prior art keywords
- service
- distance
- search
- service request
- services
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000004422 calculation algorithm Methods 0.000 title claims abstract description 36
- 238000000034 method Methods 0.000 claims description 8
- 238000012216 screening Methods 0.000 abstract description 7
- 238000010845 search algorithm Methods 0.000 abstract description 3
- 230000006870 function Effects 0.000 description 24
- 238000013519 translation Methods 0.000 description 16
- 238000010586 diagram Methods 0.000 description 9
- 238000004364 calculation method Methods 0.000 description 5
- 230000005484 gravity Effects 0.000 description 4
- 239000002131 composite material Substances 0.000 description 2
- 238000002474 experimental method Methods 0.000 description 2
- 239000002245 particle Substances 0.000 description 2
- 238000012545 processing Methods 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 201000004569 Blindness Diseases 0.000 description 1
- 241001269238 Data Species 0.000 description 1
- 241000646858 Salix arbusculoides Species 0.000 description 1
- 238000013473 artificial intelligence Methods 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000013500 data storage Methods 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 235000013399 edible fruits Nutrition 0.000 description 1
- 238000001914 filtration Methods 0.000 description 1
- 238000007726 management method Methods 0.000 description 1
- 239000000203 mixture Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 208000014451 palmoplantar keratoderma and congenital alopecia 2 Diseases 0.000 description 1
- 238000004904 shortening Methods 0.000 description 1
- 238000004088 simulation Methods 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L67/00—Network arrangements or protocols for supporting network services or applications
- H04L67/50—Network services
- H04L67/60—Scheduling or organising the servicing of application requests, e.g. requests for application data transmissions using the analysis and optimisation of the required network resources
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Computer Hardware Design (AREA)
- Evolutionary Computation (AREA)
- Geometry (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The present invention provides a kind of multirobots of SOA cloud platform to service border circular areas searching algorithm, it is to regard various specific function autonomous mobile robots as a kind of service, using service request person as the service search center of circle, the type and longitude and latitude of service are service search index, the shortest distance and grading parameters are optimal service screening index, and a kind of service search algorithm of optimal service is obtained by the way of search range is gradually expanded.
Description
Technical field
The present invention relates to the service fields of cloud robot, and in particular to a kind of multirobot service border circular areas searching algorithm.
Background technique
Multiple autonomous mobile robots service dispatch is to select optimal service in the case where service can be scheduled, and is being taken
It is artificial intelligence and multirobot area research that optimal service, which can quickly be pushed, in the case that enough, service density of being engaged in is very big
Important topic.Moreover, to dispatch robot work in dynamic, non-structured environment, since airborne ability is low, cost
Height can not carry out the problems such as high-performance calculation, limit multirobot development.Doctor Kuffer of Carnegie Mellon University in
The concept of " cloud robot (cloud robotics) " is put forward for the first time within 2010 in Humanoids meeting, by robotics
It is combined with cloud computing, the heavy calculating tasks such as data processing, trajectory planning, data storage, optimizing decision is unloaded to
Cloud, robot, which need to only carry various sensors and executing agency, can complete complicated movement, can reduce robot in this way
Load can make full use of cloud computing resources again.After this, the concept of cloud robot is widely used.2013, Ben Kehoe
" robot based on cloud grabs Google's object recognition engine " is proposed with Akihiro Matsukawa et al., it is integrated with one
Proprietary object recognition engine, the Dian Yunku of a Willow Garage PR2 robot and airborne color depth camera, Google
(PCL) crawl of Lai Shixian three-dimensional robot.2015, Gajan Mohanaraja et al. was based on RoboEarth and proposes one
New cloud robot platform --- the Rapyuta of kind, it is the online database of RoboEarth cloud engine, it is intended to allow robot
Inquiry database is to understand its environment construction and provide guidance;Jiang Yuanyuan of China Science & Technology University et al. proposes apery
Robot is theoretical, and the anthropomorphic robot of their team's designs has the movement of class people and with independent environment sensing and adaptively
Ability, it can pass through constantly academic environment information as AlphaGo and independently complete some outstanding tasks.
However general integrate carried out on the basis of existing robot research is faced with hardware isomerism, software isomerism
Etc. many technical problems.In order to solve isomerism and heterogeneity, Gartner Group is proposed earliest towards clothes within 1996
The framework (Service Oriented Architecture, SOA) of business, this service architecture is widely adopted later.2016,
Cai Y F et al. proposes a kind of face multilayer SOA of multi-robot Cooperation to solve the problems, such as isomerism in multi-Robot Cooperative
To service system structure.2017, Skarzynski, Kamil et al. proposed a kind of distributed robotic systems based on SOA
It common architecture and puppy parc for executing composite services and executes for monitoring and restores simple from failure
Agreement.
The studies above is mainly the explanation to cloud robot architectural framework and multirobot SOA interface layer, but not yet to reality
The multirobot service dispatch problem of border scene is introduced.
Summary of the invention
The present invention has designed and developed a kind of multirobot service border circular areas searching algorithm, and goal of the invention of the invention is logical
Cross border circular areas searching algorithm realize from big density, multi-quantity the service of specific function robot in select optimal service.
Technical solution provided by the invention are as follows:
A kind of multirobot service border circular areas searching algorithm, includes the following steps:
Step 1: obtaining the position of the service request person after service request person sending service request, and to service Shen
It please determine region of search, filter out the service type not in described search region centered on person;
Step 2: if there is required service in described search region, and if service needed for described is occupied, it jumps
Cross occupied service search until it is described needed for service terminate enter search range, if it is described needed for service it is unoccupied,
The service that selection distance is nearest and grading parameters are optimal is simultaneously stopped the search of the service;If do not had in described search region
Service, then expand the search radius in described search region needed for described;
Step 3: if search whole services in described search region, and if ISP service tune
Maximum tolerance time of the time less than the service request person is spent, then the ISP provides clothes for the service request person
Business;If the service dispatch time of ISP is not less than the maximum tolerance time of the service request person, the service
Supplier proposes service request as second service applicant again, and repeats the above steps that search the service dispatch time small
In the maximum tolerance time of the service request person, it is determined as required ISP.
Preferably, in said step 1 further include: service type is set as two-dimemsional number in described search region
Group, the service request split into three kinds of service types by minimization principle, and using the service request person as the center of circle, determination is searched
The radius of rope filters out the two-dimensional array not in border circular areas.
Preferably, in the step 3, the service dispatch time calculating process are as follows:
Or
In formula, xuFor ISP to the shortest distance of service request person, vsuFor the autonomous speed of service request person
Degree, vsvFor the autonomous speed of ISP.
Preferably, the maximum tolerance time is 90s.
Preferably, in the step 2, the distance is calculated to include the following steps:
Step 1, creation start node distance set and other adjacent node distance sets, and point set is created, initially
Turn to only start node;
The shortest distance of step 2, the calculating start node to other nodes, and the shortest distance is recorded in the starting
In nodal distance set, calculates the distance between other nodes node adjacent thereto and be recorded in other described adjacent node distance sets
In conjunction;
Step 3 obtains the corresponding node of the shortest distance in other described adjacent node distance sets, will be added to the point
In set;
Step 4, using the node obtained in the step 3 as intermediate point, obtain the intermediate point to adjacent node distance, such as
Fruit adjacent node is greater than the intermediate point to initial point distance and the intermediate point to the adjacent node to initial point distance
Distance and, then the distance value in the start node distance set is updated;
Step 5 repeats the step 3, the step 4, until all nodes are included in the point set.
Preferably, the distance function is calculated are as follows:
In formula, α=0.1.
Preferably, in the step 2, the grading parameters calculating process are as follows:
Y=wscore×Rscore+wsuccess×Usuccess;
In formula, wscore、wsuccessFor the weighted value of parameters, and wscore+wsuccess=1.
Preferably, the function of the grading parameters are as follows:
In formula, β=5.5.
The present invention compared with prior art possessed by the utility model has the advantages that
1, the present invention mainly has studied a kind of multirobot service border circular areas searching algorithm (CASA) of SOA cloud platform,
CASA is to regard various specific function autonomous mobile robots as a kind of service, using service request person as the service search center of circle, clothes
The type and longitude and latitude of business are service search index, and the shortest distance and grading parameters are optimal service screening index, using gradually
The mode for expanding search range obtains a kind of service search algorithm of optimal service;
2, the present invention has been built using Django as Web frame by SOA service model and interface layer, cloud platform basis
The basis the SOA cloud platform of the multirobot composite services of layer, Service Source layer composition;Using CASS in the existing registration of MySQL
Optimal service screening is carried out in service, and minimum distance calculation and shortest path are carried out in Amap using dijkstra's algorithm
Diameter planning, it is most short for target with service provision time, optimal service is provided for service request person, while to the more of SOA cloud platform
Robot service CASA working performance has carried out emulation testing;
3, the present invention in using apply self-driving travel service as example, to automobile, chat, translation service robot service dispatch
It is emulated, from simulation result it can be seen that relative to global search, the CASA of multirobot service cloud platform avoids searching for
Blindness, reduce the time of search service, have practicability;And compared with greedy algorithm, CASA is in the service search time
On reduce about 58%, have high efficiency.
Detailed description of the invention
Fig. 1 is cloud platform master-plan schematic diagram of the present invention.
Fig. 2 is border circular areas searching algorithm flow diagram of the present invention.
Fig. 3 is distance function image schematic diagram of the present invention.
Fig. 4 is grading parameters functional image schematic diagram of the present invention.
Fig. 5 is service parameter screening process figure of the present invention.
Fig. 6 meets the requirements service schematic diagram to be of the present invention.
Fig. 7 is that applicant A border circular areas of the present invention screens schematic diagram.
Fig. 8 is border circular areas screening process figure of the present invention.
Fig. 9 is optimal service track route schematic diagram of the present invention.
Figure 10 a quantity of service schematic diagram of the present invention.
Figure 10 b quantity of service schematic diagram of the present invention.
Specific embodiment
Present invention will be described in further detail below with reference to the accompanying drawings, to enable those skilled in the art referring to specification text
Word can be implemented accordingly.
As shown in Figure 1, the SOA multirobot service cloud platform framework that the present invention is built always is divided into three layers: SOA service mould
Type and interface layer, cloud platform basal layer, robot Service Source layer.
As shown in Fig. 2, a kind of multirobot service border circular areas searching algorithm (CASA) provided by the invention, CASA be by
Various specific function autonomous mobile robots regard a kind of service as, using service request person as the service search center of circle, the type of service
It is service search index with longitude and latitude, the shortest distance and grading parameters are optimal service screening index, are searched for using being gradually expanded
The mode of range obtains a kind of service search algorithm of optimal service, if the clothes that an area has thousands of kinds of robots to provide
Business, has when service request every time and (the same time may have multiple requests), how from a large amount of service the inside to select optimal clothes
Business is pushed to applicant, and this is this paper problem to be solved, is specifically comprised the following steps:
Step 1: obtaining the position of the service request person after service request person sending service request, and to service Shen
It please determine region of search, filter out the service type not in described search region centered on person;
Step 2: if having in described search region it is described needed for service, and if it is described needed for service it is occupied,
Occupied service search is then skipped until service needed for described terminates to enter search range, if service needed for described is not occupied
With the service that selection distance is nearest and grading parameters are optimal is simultaneously stopped the search of the service;If in described search region
There is no the required service, then expands the search radius in described search region;
Step 3: if search whole services in described search region, and if ISP service tune
Maximum tolerance time of the time less than the service request person is spent, then the ISP provides clothes for the service request person
Business;If the service dispatch time of ISP is not less than the maximum tolerance time of the service request person, the service
Supplier proposes service request as second service applicant again, and repeats the above steps that search the service dispatch time small
In the maximum tolerance time of the service request person.
In the present invention, the selection distance higher service of smaller and grading parameters, and if adjust the distance with grading parameters not
It processes, then servicing selection can be very high to the susceptibility of the two.It adjusts the distance when based on services selection and is wanted with the special of grading parameters
Ask, can choose distance function and the two is normalized in grading parameters function, select optimal service function be in order to
Indicate distance function and grading parameters the functional value specific gravity shared by different distance ranges.
The selection principle of destination service: the selection distance higher service of smaller and grading parameters, and if adjust the distance and comment
Point parameter is not processed, then servicing selection can be very high to the susceptibility of the two.It adjusts the distance when based on services selection and grading parameters
Particular/special requirement, can adjust the distance and be normalized with grading parameters.
As shown in figure 3, distance function selects: when services selection the chance that is selected to more greatly of distance it is smaller, apart from smaller quilt
It is bigger to choose probability;It selects Gauss attenuation function for distance function in the present invention, chooses α=0.1;
With the increase of distance x between user and service, user selects the probability f (x) of the service to decrease, and works as distance
When constantly increasing, i.e., the degree of correlation between service and user goes to zero.
As shown in figure 4, grading parameters function selects: with distance function on the contrary, the probability that is selected to more greatly of grading parameters more
Greatly;It is as follows that grading parameters function is chosen herein:
In formula, β=5.5 are chosen.
When grading parameters y is larger, grading parameters function g (y) speedup is larger, and the superior degree difference of services selection is larger, instead
Services selection superior degree difference it is smaller, meet service and choose and require.
Optimal service function determines: choosing F (x, y)=λ f (x)+η g (y) and is used as optimal service Selection of Function, λ, η in formula
For f (x), g (y) than column coefficient, i.e. the specific gravity shared by different region of search the two, value gropes to acquire by experiment.It is comprehensive
On, optimal service function is as follows:
When distance is in different range, different proportionality coefficient λ, η are chosen, distance function f (x) is shared when 0≤x≤2.5
Specific gravity it is larger, specific gravity shared by scoring parametric function g (y) is larger when 2.5 < x≤4.5.Finally, with the maximum value of F (x, y)
For optimal objective.
In another embodiment, in step 2, shortest distance meter is carried out using dijkstra's algorithm and Amap
Calculation and shortest path planning.It is that starting point reaches target section that dijkstra's algorithm, which is with some node (ISP) of Amap,
The distance value and the smallest algorithm on side between point (service request person) passed through node.Node needed for dijkstra's algorithm, road
Amap API is called to obtain by Python crawlers with range data.Dijkstra's algorithm is in shortest distance meter
Need to state three set: start node distance set BDIS, other adjacent node distance set ODIS and point set when calculation
PIO.Start node distance set BDIS stores the shortest distance between start node and other nodes, when two nodes are non-conterminous
When, distance is ∞;Other adjacent node distance sets ODIS stores the distance between other nodes and its phase adjacent node;Point set
Close the node that PIO storage has acquired the shortest distance.Assuming that start node is bdisa, destination node bdisb, dijkstra's algorithm
Include:
Step 1, creation start node distance set BDIS, are initialized as sky;
Step 2 creates other adjacent nodes distance set ODIS, is initialized as sky;
Step 3, creation point set PIO, are initialized as only start node bdisa;
Step 4, calculate node bdisaIt is recorded in BDIS to the shortest distance of other nodes, and by the shortest distance, so that
BDIS={ ' bdis1': da1..., ' bdism': inf ..., ' bdisb': dab(bdis indicates that node, d arrive for previous node
Distance between latter node, inf indicate that distance value is infinitely great);Calculate the distance and record between other nodes node adjacent thereto
In ODIS, so that ODIS={ ' bdisI': { ' bdis2': dI2..., bdisI': dIl..., ' bdisk': ' { ' bdiso':
dko..., ' bdisp': dkp(wherein);
Step 5 obtains the corresponding node bdis of the shortest distance in BDISn, by bdisnIt is added in PIO;
Step 6, with bdisnFor intermediate point, bdis is obtainednTo adjacent node (not including the node in set PIO) distance,
If BDIS [bdish] > BDIS [bdisn]+ODIS[bdisn][bdish];
(bdishIt is some and bdisnAdjacent node), then update bdis in BDIShThe distance value of node, so that dh=dn
+dnh;
Step 7 repeats step 5,6, until all nodes are included in set PIO.
To sum up, node bdisaTo node bdisbShortest distance x=BDIS [bdisb], shortest path is PIO interior joint
Sequence.
Meanwhile in the present embodiment, for the reliability of algorithm, now do it is assumed hereinafter that:
1, there are many service of service request person region and service closeness is larger, i.e., the energy after limited times service search
Positioning service;
2, service request person region network signal is good;
3, service request person region traffic convenience;
4, type, longitude and latitude, the service-number etc. serviced is all previously stored in lane database, and fills above each service
There are GPRS locator and networked devices, the service end position of oneself can be uploaded after service.
In another embodiment, the selection principle of this service grading parameters includes two aspects: service request person's
Service satisfaction scoring, service dispatch success rate index.
1, the service satisfaction scoring R of service request personscoreThe person that is service request proposes each service after service
Scoring of the donor in this service;
2, service dispatch success rate index UsuccessIt is that ISP has been scheduled to including this service
The probability of function, success rate index are affected to optimal service function, can obtain grading parameters y in summary are as follows:
Y=wscore×Rscore+wsuccess×Usuccess; (4)
In formula, wscore、wsuccessFor the weighted value of parameters, and werocs sseccus+w=1;As a kind of excellent
Choosing, in the present embodiment, takes wscore=0.4, wsuccess=0.6;It obtains after this service grading parameters and last scoring
Parameter, which is averaged, can obtain the scheduled grading parameters of service waiting.
In another embodiment, border circular areas service screening is carried out in step 1~step 3 to specifically include:
Step 1, we every kind of service all regarded as a two-dimensional array [Stype, Llon-lat], wherein StypeIndicate service
Type, Llon-latExpression service type is StypeCorresponding longitude and latitude;In this example, it is assumed that lane database has many services,
It is expressed as the array list of table 1 according to the form of two-dimensional array;
1 S-L array list of table
(longitude and latitude is (Lon by step 2, certain service request person0, Lat0)) application father service pass through Logic minimization principle
Splitting into service type is Sk-1、Sk、Sk+1Three kinds of sub-services (wherein, father's service can need to select a variety of according to oneself);
Firstly, it is S that cloud platform, which selects service type,k-1、Sk、Sk+1Two-dimensional array, such as the cell of the 4th~6 row in table 1;
Wherein, Logic minimization principle refers to service management center upon receipt of a service request, according to service distribution
Father's service (block service) is split into the sub-services of multiple logic states by the maximum principle of minimum number, service availability, different
Sub-services there are multiple service entities for being scattered in different zones with the sub-services type to be corresponding to it again.
Step 3, using service request person as the center of circle, R be radius determine border circular areas, filter out two not in border circular areas
Dimension group for example obtains the array [S in border circular areask-1, L2]、[Sk-1, Ls]、[Sk-1, Ll] (wherein 1≤s, l≤m);
Step 4, the two-dimensional array fallen in border circular areas obtained by step 3, take the L of -1 row of kth respectivelyμ(μ=2, s,
L) it is calculated by dijkstra's algorithm to applicant's longitude and latitude (Lon0, Lat0) the shortest distance and plan shortest path;
For Sk-1There are three types of longitude and latitude L2(LonL2, LatL2)、Ls(LonLs, LatLs)、Ll(LonLl, LatLl), respectively
Obtain shortest distance x1、x2、x3, the corresponding grading parameters of three are y1、y2、y3, obtain three groups of solution (x1, y1)、(x2, y2)、(x3,
y3), it brings formula (3) into respectively and obtains grading parameters F1、F2、F3.If F1> F2> F3, then F can be determined1Corresponding service type
Sk-1Longitude and latitude be L2(LonL2, LatL2), while being L by longitude and latitude2(LonL2, LatL2) corresponding with service shortest path letter
Breath is transferred to ISP parsing, then mobile to service request person, and stopping service type at this time is Sk-1Search;
Step 5, due to service type SkAnd Sk+1Not in the border circular areas that radius is R, so on the basis of former radius
Increase n Δ R (n >=1), existing border circular areas radius R+n Δ R (n >=1) repeats step (2), (3), (4), finds service type Sk
And Sk+1Longitude and latitude Lh(LonLh, LatLh)、Lj(LonLj, LatLj) (wherein 1≤h, j≤m);
Step 6 obtains service type Sk-1、Sk、Sk+1Longitude and latitude L2(LonL2, LatL2)、Lh(LonLh, LatLh)、Lj
(LonLj, LatLj) after, by the interface message of these three services and the walking path person that is sent to service request;
Step 7, due to many service robots movement speed more slowly and when service dispatch distance farther out when (such as
The S of service request person applicationkService and Sk+1Service), the service dispatch time is increased, the round-robin scheduling rate of service is reduced.For
Shortening service dispatch time, certain ISP SuAccording to arrive service request person RuShortest distance xuWith autonomous speed
vsuThere can be service dispatch time TsuAnd Tsu' calculate following (v in formula (6)svFor SuThe service autonomous movement speed of application):
If service dispatch time TsuMaximum greater than people endures time T=90s, then is used as service request person RvIt (is aforementioned
ISP Su) repeat step 2~6 other service S of schedulingv(autonomous speed is vsv);If Tsu′+Tsv≥Tsu(Tsv
For ISP SvThe service dispatch time), then abandon dispatch service Sv, itself is used as ISP SuReach service Shen
It please person Ru。
Embodiment
In the present embodiment, SOA multirobot service cloud platform selects notebook AMD A6-5350M processor, in 8GB
It deposits, AMD Radeon HD (8450G+8570M) double video cards.Cloud platform part is using Nginx as cloud reverse proxy service
Device realizes parallel computation in conjunction with Lua, and Django is as Web frame, and MySQL is as back-stage management database, PyCharm conduct
Programmable device, Python is as programming language, and using Restful software architecture style, experiment is from Amap official with longitude and latitude
What net obtained.
As shown in figure 5, there is 30000 service datas in database at present, (longitude and latitude is service request person
[117.2390864,39.2209323]) self-driving travel service is needed, self-driving travel service is torn open according to Logic minimization principle first
It include automobile services (there is particle carrying capacity), chatting service, three kinds of translation service services after point, then execution order filtering goes out full
Longitude, dimension, the service-number for the service that these three service types of foot require, finally by all service numbers for meeting service request
According to being packaged into JSON format transmission to front end.
As shown in fig. 6, the service point for meeting service type requirement is shown in Amap, wherein circle indicate with A be
The range searching circle of the heart.
As shown in Figure 7, Figure 8, since the service point that meets the requirements is too many and apart from too far, so first determining that radius is 1km
Search circle, be shown in the service point that meets the requirements in circle, filter out the service point outside circle, calculate separately each service point to taking
The shortest distance x of business applicant, and the shortest distance is assigned to the corresponding service of the service point, if without full within the scope of 1km
It requires enough, then search range is gradually expanded, until searching all service type data.
The service point filtered out by border circular areas carries out the calculating of optimal service functional value, the selection result such as table in cloud platform
2。
Service the selection result within the scope of 1000 meters of 2 service request person of table
Can be obtained from upper face data meet in automobile services condition service longitude and latitude be [117.2350593,
39.2249956], calculating optimal service functional value by formula (3) is 0.955208;Similarly, the optimal of condition is met in chatting service
Service longitude and latitude is [117.2340694,39.2192569], and optimal service functional value is 0.876785.
Due to not searching translation service in the range of 1km, so increasing search radius, (R=R+ Δ R, Δ R is each
Increase 0.5km), when search radius be 1.5km, such as Fig. 7, obtain the selection result such as table 3 of translation service.
Service the selection result within the scope of 3 service request person 1.5km of table
As automobile services and chatting service, the optimal service of translation service learn from else's experience latitude be [117.2422294,
39.2316456], optimal service functional value be 0.744387.
As shown in figure 9, after obtaining the longitude and latitude of optimal service, by the essential information, interface parameters, shortest path of optimal service
Diameter information is sent to service request person, 1 is service request person in Fig. 9, and 2,3,4 be respectively automobile services, chatting service, translation
Service.
The shortest distance x difference of automobile services, chatting service, translation service to service request person is obtained by Amap
For 0.432306km, 0.418km, 1.170km, it is assumed that the autonomous speed of three is respectively 5m/s, 0.5m/s, 0.5m/s,
The service dispatch time T of three is obtained by formula (5)sRespectively 86.46s, 836s, 2340s, it is known that chatting service and translation service
Service dispatch time TsGreater than 90s, and automobile services have faster autonomous speed and particle carrying capacity, so in order to shorten
Service dispatch time, chatting service and translation service as service request person (longitude and latitude be respectively [117.2340694,
39.2192569], [117.2422294,39.2316456]) application automobile services.
1, the selection result of chatting service such as table 4.
4 chatting service of table applies for the result of automobile services as service request person
Latitude of learning from else's experience is [117.2350694,39.2212569], and the automobile that optimal service functional value is 0.974561 takes
Business, the shortest distance x to chatting service are 0.232km, and autonomous speed is 5m/s, obtain the service dispatch time by formula (5)
TsFinal destination is reached after having travelled 0.418km after chatting service in load after reaching chatting service location for 46.4s, is used
When 83.6s, so chatting service reach final destination the final used time be 130s, relatively application service before save the used time
706s。
2, the selection result of translation service such as table 5.
The result of automobile services is applied in 5 translation service of table as service request person
Latitude of learning from else's experience is [117.2431652,39.2331326], and the automobile that optimal service functional value is 0.986571 takes
Business, the shortest distance x to translation service are 0.257km, and autonomous speed is 5m/s, obtain the service dispatch time by formula (5)
TsFor 51.4s.After reaching translation service location, final destination, used time are reached after having travelled 1170m after translation service in load
234s.If after automobile services carry upper translation service, included navigation system breaks down when driving to half, original place application is needed to lead
Boat service, the selection result such as table 6.
The result of navigation Service is applied in 6 automobile services of table as service request person
Latitude of learning from else's experience is [117.2435642,39.2343565], and the navigation that optimal service functional value is 0.984271 takes
Business, the shortest distance x to automobile services are 0.166km, it is assumed that autonomous speed is 1.5m/s, must service tune by formula (5)
Spend time TsFor 110.67s, if skimming service request and data processing time, the row of calculation ISP to service requester
It sails the time, then translation service reaches the time of final service request person as 279.07s, and relatively application services the saving used time before
2060.93s。
As shown in Figure 10 a, Figure 10 b, by above-mentioned algorithm, it is assumed that service request quantity is three, existing quantity of service 0~
In 30000 ranges, CASA and common greedy algorithm (Greedy Algorithm) are subjected to the comparison of service search time, schemed
In 10a, as the quantity of registration service is increasing, the trend constantly risen is presented in service search time, works as service registration
When quantity is 10000, CASA saves used time nearly 53%, when service registration quantity is continuously increased 30000, the two compared with greedy algorithm
Although the service search time all increased, CASA averagely saves the used time compared with greedy algorithm also greater than 50%;Similarly, it is assumed that
Existing quantity of service is 30000, and service request quantity carries out the service search time in 0~8 variation, by CASA and greedy algorithm
Comparison, in figure 10b, with being increasing for service request quantity, the trend constantly risen is also presented in the service search time, works as clothes
For business application quantity when reaching 8, CASA averagely saves the used time nearly 63% compared with greedy algorithm, therefore, CASA proposed by the present invention compared with
Waiting time of the greedy algorithm when dispatching optimal service is relatively small, and the real-time of cloud platform service is preferable.
Although the embodiments of the present invention have been disclosed as above, but its is not only in the description and the implementation listed
With it can be fully applied to various fields suitable for the present invention, for those skilled in the art, can be easily
Realize other modification, therefore without departing from the general concept defined in the claims and the equivalent scope, the present invention is simultaneously unlimited
In specific details and legend shown and described herein.
Claims (8)
1. a kind of multirobot services border circular areas searching algorithm, which comprises the steps of:
Step 1: obtaining the position of the service request person after service request person sending service request, and with service request person
Centered on, it determines region of search, filters out the service type not in described search region;
Step 2: if there is required service in described search region, and if service needed for described is occupied, skip by
Occupying service search service needed for described terminates to enter search range, if servicing unoccupied, selection needed for described
The service that distance is nearest and grading parameters are optimal is simultaneously stopped the search of the service;If without described in described search region
Required service then expands the search radius in described search region;
Step 3: if search whole services in described search region, and if when the service dispatch of ISP
Between be less than the service request person the maximum tolerance time, then the ISP provides service for the service request person;
If the service dispatch time of ISP is not less than the maximum tolerance time of the service request person, the service is provided
Person proposes service request as second service applicant again, and repeats the above steps and search the service dispatch time less than institute
The maximum tolerance time for the person that states service request, it is determined as required ISP.
2. multirobot as described in claim 1 services border circular areas searching algorithm, which is characterized in that in said step 1
Further include: service type is set as two-dimensional array in described search region, the service request is torn open by minimization principle
It is divided into three kinds of service types, using the service request person as the center of circle, determines the radius of search, filter out not in border circular areas
Interior two-dimensional array.
3. multirobot as described in claim 1 services border circular areas searching algorithm, which is characterized in that in the step 3
In, the service dispatch time calculating process are as follows:
Or
In formula, xuFor ISP to the shortest distance of service request person, vsuFor the autonomous speed of service request person, vsv
For the autonomous speed of ISP.
4. multirobot as claimed in claim 3 services border circular areas searching algorithm, which is characterized in that when the maximum tolerance
Between be 90s.
5. multirobot as described in claim 1 services border circular areas searching algorithm, which is characterized in that in the step 2
In, it calculates the distance and includes the following steps:
Step 1, creation start node distance set and other adjacent node distance sets, and point set is created, it is initialized as
Only start node;
The shortest distance of step 2, the calculating start node to other nodes, and the shortest distance is recorded in the start node
In distance set, calculates the distance between other nodes node adjacent thereto and be recorded in other described adjacent node distance sets
In;
Step 3 obtains the corresponding node of the shortest distance in other described adjacent node distance sets, will be added to the point set
In;
Step 4, using the node obtained in the step 3 as intermediate point, obtain the intermediate point to adjacent node distance, if phase
Neighbors to initial point distance be greater than the intermediate point to initial point distance at a distance from the intermediate point to the adjacent node
With then the distance value in the start node distance set is updated;
Step 5 repeats the step 3, the step 4, until all nodes are included in the point set.
6. multirobot as claimed in claim 5 services border circular areas searching algorithm, which is characterized in that calculate described apart from letter
Number are as follows:
In formula, α=0.1.
7. multirobot as claimed in claim 6 services border circular areas searching algorithm, which is characterized in that in the step 2
In, the grading parameters calculating process are as follows:
Y=wscore×Rscore+wsuccess×Usuccess;
In formula, wscore、wsuccessFor the weighted value of parameters, and wscore+wsuccess=1.
8. multirobot as claimed in claim 7 services border circular areas searching algorithm, which is characterized in that the grading parameters
Function are as follows:
In formula, β=5.5.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811492950.5A CN109670224B (en) | 2018-12-07 | 2018-12-07 | Multi-robot service circular area searching algorithm |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811492950.5A CN109670224B (en) | 2018-12-07 | 2018-12-07 | Multi-robot service circular area searching algorithm |
Publications (2)
Publication Number | Publication Date |
---|---|
CN109670224A true CN109670224A (en) | 2019-04-23 |
CN109670224B CN109670224B (en) | 2023-11-21 |
Family
ID=66144724
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201811492950.5A Active CN109670224B (en) | 2018-12-07 | 2018-12-07 | Multi-robot service circular area searching algorithm |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109670224B (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN114506596A (en) * | 2022-03-18 | 2022-05-17 | 北京云迹科技股份有限公司 | Method and device for calling intelligent mobile collection equipment |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN204215204U (en) * | 2014-11-19 | 2015-03-18 | 济南大学 | A kind of scenic spot intellect service robot |
CN105844364A (en) * | 2016-04-08 | 2016-08-10 | 上海派毅智能科技有限公司 | Service robot optimal path program method based on heuristic function |
CN106448138A (en) * | 2016-11-09 | 2017-02-22 | 中国科学技术大学苏州研究院 | Optimal multi-vehicle scheduling method based on active distribution type taxi service system |
CN107545486A (en) * | 2016-06-28 | 2018-01-05 | 滴滴(中国)科技有限公司 | One kind uses car order processing method and server |
CN107608349A (en) * | 2017-09-11 | 2018-01-19 | 北京九星智元科技有限公司 | Scenic spot service robot system |
US20180333847A1 (en) * | 2016-01-04 | 2018-11-22 | Hangzhou Yameilijia Technology Co., Ltd. | Method and apparatus for working-place backflow of robots |
-
2018
- 2018-12-07 CN CN201811492950.5A patent/CN109670224B/en active Active
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN204215204U (en) * | 2014-11-19 | 2015-03-18 | 济南大学 | A kind of scenic spot intellect service robot |
US20180333847A1 (en) * | 2016-01-04 | 2018-11-22 | Hangzhou Yameilijia Technology Co., Ltd. | Method and apparatus for working-place backflow of robots |
CN105844364A (en) * | 2016-04-08 | 2016-08-10 | 上海派毅智能科技有限公司 | Service robot optimal path program method based on heuristic function |
CN107545486A (en) * | 2016-06-28 | 2018-01-05 | 滴滴(中国)科技有限公司 | One kind uses car order processing method and server |
CN106448138A (en) * | 2016-11-09 | 2017-02-22 | 中国科学技术大学苏州研究院 | Optimal multi-vehicle scheduling method based on active distribution type taxi service system |
CN107608349A (en) * | 2017-09-11 | 2018-01-19 | 北京九星智元科技有限公司 | Scenic spot service robot system |
Non-Patent Citations (2)
Title |
---|
张加一: "游戏AI中的路径搜索算法的研究与应用", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
蔡云飞: "多机器人协作定位及***架构研究", 《中国博士学位论文全文数据库 信息科技辑》 * |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN114506596A (en) * | 2022-03-18 | 2022-05-17 | 北京云迹科技股份有限公司 | Method and device for calling intelligent mobile collection equipment |
Also Published As
Publication number | Publication date |
---|---|
CN109670224B (en) | 2023-11-21 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US11231715B2 (en) | Method and system for controlling a vehicle | |
Loquercio et al. | Deep drone racing: From simulation to reality with domain randomization | |
Wang et al. | A reinforcement learning approach for UAV target searching and tracking | |
Choi et al. | Unmanned aerial vehicles using machine learning for autonomous flight; state-of-the-art | |
González-Banos et al. | Navigation strategies for exploring indoor environments | |
Soltero et al. | Decentralized path planning for coverage tasks using gradient descent adaptive control | |
Xiang et al. | Continuous control with deep reinforcement learning for mobile robot navigation | |
Elamvazhuthi et al. | Optimal control of stochastic coverage strategies for robotic swarms | |
Jiang et al. | A brief survey: Deep reinforcement learning in mobile robot navigation | |
CN114169066A (en) | Space target characteristic measuring and reconnaissance method based on micro-nano constellation approaching reconnaissance | |
Wu et al. | UAV Path Planning Based on Multicritic‐Delayed Deep Deterministic Policy Gradient | |
CN109670224A (en) | A kind of multirobot service border circular areas searching algorithm | |
Chen et al. | Deep reinforcement learning-based robot exploration for constructing map of unknown environment | |
Ou et al. | Hybrid path planning based on adaptive visibility graph initialization and edge computing for mobile robots | |
Azoulay et al. | UAV Flocks Forming for Crowded Flight Environments. | |
Zhou et al. | Research on Circular Area Search algorithm of multi-robot service based on SOA cloud platform | |
Zhou et al. | Novel task decomposed multi-agent twin delayed deep deterministic policy gradient algorithm for multi-UAV autonomous path planning | |
Yang | Reinforcement learning for multi-robot system: A review | |
Edelkamp et al. | Monte-carlo search for prize-collecting robot motion planning with time windows, capacities, pickups, and deliveries | |
Verma et al. | Computational investigation of environment learning in guidance and navigation | |
Xu | Context-based trajectory prediction with LSTM networks | |
Kumar et al. | Longitudinal aerodynamic parameters estimation using machine learning with neuro artificial bee colony fusion algorithm (NABC) | |
Gao et al. | Developing an effective algorithm for dynamic UAV path planning with incomplete threat information | |
Dicheva et al. | 3D waypoint generation in a dynamic environment for an airborne launch mission | |
Xihan et al. | Robot in a china shop: Using reinforcement learning for location-specific navigation behaviour |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |